by L. C. Arulpragasam

In 1960 while in the Ceylon Civil Service, I was given charge of planning for the Agriculture Sector in the Department of National Planning, with the duty of looking into the economics and effects of the irrigated colonization schemes in the Dry Zone, among others.

These schemes were considered by successive governments as their major vehicle in achieving their stated objectives of achieving self-sufficiency in rice and increasing employment.

Our studies, however, showed that far from reaching these goals, the employment they created for each year of investment from 1950 to 1960 provided new employment for only six percent of the annual increase in the workforce. Likewise, far from reducing the past overall deficit in rice production, the annual production of all the new paddy lands created each year under colonization schemes was able to feed only one-tenth of the annual increase in population in that same year.

That is, far from reducing the accumulated deficit in employment and rice production, we were actually losing the race in attaining rice self-sufficiency and in attaining full employment by 90% each year (The Short Term Implementation Programme, Department of National Planning, 1962). But this was the major plank of Sri Lanka’s agricultural development policy from Independence up to 1975, absorbing by far the greatest part of our non-export agricultural development budget at that time.

In order to investigate this issue further, benefit-cost studies were undertaken in the Department of National Planning in 1960 of two irrigated colonization schemes in the Dry Zone, namely, the Mahakandarawa and Rajangana Schemes. Taking all the benefits of these schemes over 50 years on one side and all their costs over the same period on the other, the benefit: cost ratio (discounted over 50 years) was 0.56 in respect of the Mahakandarawa Scheme and 0.67 in respect of the Rajangana Scheme. Simply put, for one rupee spent in 1960 we would get back only 56 cents out of that one rupee after 50 years in respect of the Mahakandarawa scheme, and 67 cents after 50 years in respect of the Rajangana Scheme.

This was based on the prices prevailing at that time, with the yield of paddy calculated at 42 bushels an acre, which was higher than the yield of 40 bushels per acre claimed by Government.1 However, from the cost side, it was recognized that there was a heavy social welfare component built into these schemes – such as the provision of houses, schools and clinics for the colonists – which were not strictly agricultural development costs, although they were needed to put the colonists in place.

Likewise on the benefit side, there were unquantifiable benefits from colonization, such as relief of congestion in crowded areas and the economic benefits arising from population redistribution, which were not included in this calculation. The above figures and analysis were accepted by the Ministry of Agriculture and Lands and the Ministry of Finance and Planning before they were published in the “Short-Term Implementation Programme” of the Sri Lanka Government in 1962.

The benefit: cost ratios contained in the above Programme were contested in Parliament by Mr. Dudley Senanayake (Leader of the Opposition at that time) who had been the Minister of Agriculture and Lands in previous Governments. He vowed to look into this question when his party (the UNP) came back into power. This was done by his succeeding UNP Government through a Committee for the Evaluation of the Gal Oya Scheme in 1969, headed by Dr. B.H. Farmer who was known for his pioneering work: “Pioneer Peasant Colonization in Ceylon” (1957). This Committee, having investigated the Gal Oya Scheme, arrived at a benefit: cost ratio of 0.58 discounted over a 50 year period for this Scheme, as reported in Sessional Paper No.1 of 1970.

That is, after all benefits and costs are counted over a period of 50 years, the return on one rupee under the Gal Oya Scheme would be only 58 cents after 50 years, less than the average of the returns of the two schemes studied by me, which had been challenged by Mr. Dudley Senanayake in Parliament.

A few months earlier (in 1960), because my findings were highly critical of the central plank of the Government’s development programme, I was taken by the Minister of Finance and Planning (Mr. Felix Dias Bandaranaike) to the Hon. Prime Minister, Mrs. Sirimavo Bandaranaike. She came straight to the point, saying more or less the following: “Mr. Arulpragasam, I understand that you are saying that our colonization programme, our main development programme, is wasteful and uneconomic. How can this be?”

My position was actually, first, that if the aim was to increase rice production for self-sufficiency, this could be faster and more economically achieved by agricultural intensification (increasing yields and cropping intensities) rather than through the very slow, costly and wasteful programme of land development and colonization being carried out at that time.

Secondly, my argument was not against the colonization schemes per se, because they were undoubtedly economically necessary and desirable for the country over the longer term.

My point was that these schemes could never realize their stated objective if they were planned and implemented in their then wasteful, uneconomic way. Mrs. Bandaranaike had only recently become Prime Minister and did not know much about irrigation, colonization or agriculture at that time. So I had to summarize my conclusions as simply and forcefully as I could. I explained it in the following terms. “Madam, this reminds me of our University exams, where we were given five questions to answer, with 20 marks per question. This is like what a colonist gets: a 5 acre allotment (in 1960) made up of two acres of highland and three acres of paddy land. According to the starting premise of the colonization schemes, the lands in the dry zone cannot be cultivated without irrigation.

If that be the case, the two acres of highland (which are not irrigated) cannot be cultivated because of lack of water.” In fact, at that time not even 8% of all the lands in highland allotments in the colonization schemes were actually cultivated. “This means, Madam, I cannot answer two questions (the two acres of highland accounting for 40 marks in my hypothetical exam at 20 marks per question) which leaves me with a potential maximum of only 60 marks. But I still have three acres of irrigated paddy land which are supposed to be cultivated with two crops of paddy per year.

However in practice, only one paddy crop is actually cultivated under most of these schemes”. In fact in 1960, not even 10% of all paddy land under irrigated colonization schemes was being cultivated for two seasons. “This means, Madam, that I can get a maximum of only 30 marks out of my potential 100 marks – which means that I have failed my exam before even putting pen to paper!”

She understood immediately, and asked the obvious question: “But why is this so?”

To sum up, there was no way in which these schemes could be made economically viable unless and until: (i) The yield of paddy on these lands could be doubled, which was shown to be feasible at that time (ref. The Short-Term Implementation Programme,1962); (ii) All paddy lands under these schemes were cropped twice a year, which was also feasible at that time (ibid). (iii) The highland allotments were brought fully under production with high value crops. (iv) The cost of irrigation (the most costly item) could be drastically reduced, which could also be feasibly done (ibid).

General Considerations

The starting premise in the policy and planning of these irrigated colonization schemes was that stable and profitable cultivation in the Dry Zone was not possible without irrigation. Frankly, I did not accept even this starting premise. Alternatives had already been explored by that time (1960) through the pioneering work of Dr. Ernest Abeyratne at the Maha Illupalama Dry Zone Research Institute, although not always for paddy cultivation. But since the Government had always insisted on rice cultivation as the basis for peasant colonization, irrigation was admittedly needed for the second rice crop.

Much of the irrigation provided to the major colonization schemes was based on the restoration and augmentation of ancient tanks or reservoirs. But the safety factor adopted by the Irrigation Department at that time (1960) often erred too much on the side of safety. The Irrigation Adviser to the National Planning Commission in India, whom the Department of National Planning brought in to advise specifically on this issue, recommended that in many of the irrigation reservoirs reviewed by him in Sri Lanka, the bund, sluice and spill levels could be raised to store more water while still being able to meet adequate safety standards, which had already been accepted in India. But the Director of the Irrigation Department did not accept this recommendation at that time (1960).

Obviously there have been changes since then. For example, I read in a Sri Lanka newspaper in February of 2010 that the Government had decided to raise the bund, sluice and spill levels of the Unnichai Irrigation reservoir in the Batticaloa District. I was amused because when I had made the same request 50 years earlier, I was told by the Irrigation Department that it was not technically advisable! Suffice it to say that if such expanded irrigation could be diverted to the highland allotments for less water-consuming but higher-value crops, the economics of these schemes could be drastically improved.

The next phase of planning relates to the level at which the irrigation channel is to be run. But this decision was based on the Irrigation Department’s assumption of gravitational irrigation only – since the Irrigation Department refused to consider anything else (an exception was the Rajangana Scheme). This meant that all lands above the channel (the so-called “highlands”) were rendered non-irrigable, virtually consigning them to non-cultivation.

One was tempted to ask (as I did, tongue in cheek) why these lands are called “highlands” and why the others are called “paddy lands”. The exasperated reply from the Director of Irrigation was: “Can’t you understand? The highlands are high because they are above the channel, and the paddy lands are paddy lands because they are below the channel!” In other words, the Irrigation Department’s own actions decided whether the lands were high or low!

It was illogical, to say the least. First, if the land above the channel is called “highland” (although it is not high except in relation to a channel drawn by the Irrigation Department), then the land below the channel should be called “lowland”. But to label it as “paddy land” was not to make an elevation judgment but an agronomic one, namely, that the land was suitable for paddy cultivation, even if it were not. Secondly, land use was being determined by the arbitrary level of a man-made channel.

If for example, a higher level channel could be contrived by raising the bund and sluice levels by three feet, or if the water could be pumped into a channel three feet higher, would these “highlands” immediately become “paddy lands”? Thirdly, even if the lands below the channel were not suitable for paddy cultivation, would they still be “paddy lands?” Needless to say, a “paddy land” should be judged by its soil and agronomic suitability for paddy cultivation and not by its elevation in relation to an arbitrary man-made channel.

This division of highland from paddy land based on the level of the channel has had three adverse consequences on agricultural development in Sri Lanka. First, the assumption of gravitational irrigation, all but renders the two acres of “highland” uncultivable.

Secondly, it relegates all land below the channel to paddy cultivation – even if it were not suitable for paddy. But thirdly, this kills any hope of crop diversification (a goal stressed in all our development plans) by ensuring that all irrigation in the dry zone is devoted only to paddy cultivation.

A key land and water use problem is how to make the so-called “highlands” more productive, since there is no way in which these schemes can be made economic when they leave 40% of their lands almost uncultivated. If this cannot be achieved by augmenting the tank and running the channel at a higher level, lift-irrigation needs to be considered. This could theoretically be done by lifting water to a higher- level channel which would then make the “highlands” irrigable. Obviously such lift-irrigation can be economic only if utilized for less water-using, higher-value crops than paddy.

When I enquired in writing about the possibility of lift-irrigation, the Director of Irrigation replied to the Director of National Planning (in 1960) in these astounding words: “Dear Sir: This is to inform you that lift irrigation is uneconomic in Sri Lanka .”- Sgd. Director of Irrigation! Thus lift-irrigation is economic in all other countries, but not in Sri Lanka! Needless to say, the Director of Irrigation was thinking of irrigation for paddy only, and that too at the outrageous rates of duty (water consumption) that prevailed in the colonization schemes at that time. For example, as shown in the Short-Term Implementation Programme (1962), chillies or onions could be grown on the highlands, yielding more than three times the returns from paddy at that time but using less than one-third the water that was being used for paddy.

The Mahaweli Scheme was later able to show the feasbility of other profitable crops on the so called paddy lands. Even very limited irrigation of the highlands for tree crops, using less than one-fifteenth of the water then being used for paddy, could probably have provided economic returns.

A foundational error in agricultural policy in 1961 was the belief that the additional rice production needed for self-sufficiency could best and most economically be obtained by expanding the area under cultivation through land development (agricultural extensification).

The possibility of increasing yields (agricultural intensification) was never considered by the Ministry of Agriculture and Lands at that time. This was like “playing Hamlet (agricultural production) without the Prince of Denmark (agriculture)”, as I argued at that time. The paddy yield of 40 bushels per acre claimed by Government for the Maha season was no better than the average yields achieved in the rainfed paddy lands in the Dry Zone, which were obtained without any of the investments made in irrigation or colonization. (ibid), while the yields in the Kandy and Kegalle districts at that time were 50-55 bushels per acre, without any of the above investment.

As pointed out at that time, these colony lands should have been capable of providing the highest yields of all paddy lands in the country because of their assured irrigation. Meanwhile, all the Cabinet Papers from the Ministry of Agriculture and Lands (up to 1965) requested more funds in order to clear more lands so as to obtain this same low yield of 40 bushels an acre, which automatically made these schemes uneconomic.

As opposed to this policy of agricultural extensification, the Short-Term Implementation Programme of 1962 (based on the H4 paddy variety already developed by our agricultural researchers by 1958) proposed a policy of agricultural intensification, arguing that even a small increase in yields of even four bushels per acre on all paddy lands outside the colonization schemes would result in more paddy production per year than all the paddy produced per year from all the existing colonization schemes put together – at almost no additional cost!

In fact, at that time, on orders of the Ministry, agricultural extension workers from the Department of Agriculture were not allowed to work in the colonization schemes, this task being left to the Land Development Officers (LDOs), with their emphasis on land development and checking violation of LDO permits: again playing “Hamlet without the Prince of Denmark”!

Hence there were many factors contributing to the poor economics of the colonization schemes in 1960. Since irrigated colonization schemes were not bad or uneconomic per se, it is necessary to look into shortcomings in planning and implementation, as a means of success in the future.

Recommendations

Irrigation

1. Possibilities of Augmentation

Although the possibility of augmenting existing reservoirs (by raising their sluice, spill and bund levels) has been done on a case-by-case basis, this needs to be done on a systematic basis, whereby all tanks/reservoirs are re-examined from this point of view. For it is well known that – other things being equal – the augmentation of existing tanks is much more cost-effective per acre foot of water delivered than its provision through brand new schemes. The safety standards themselves need to be reviewed with reference to standards adopted in other countries – which has hopefully already been done.

2. Bringing Down the Cost of Irrigation

The Short-Term implementation Programme came to the following conclusions regarding the unnecessarily high cost of irrigation, all of which were ultimately accepted by the Director of Irrigation and the Minister of Agriculture and Lands in front of the Prime Minister in 1961/62.

(i) Almost no major irrigation scheme was finished on time. The Director of Irrigation admitted in front of the Prime Minister that the Rajangana Scheme which had taken 18 years to complete could have been finished in eight years (Ibid, p. 125). He admitted that the reason for this was that too many works had been undertaken at the same time, such that none could be finished in time. This led (among others) to a dramatic fall in allotments given out to colonists, reaching the abysmally low figure of only 621 allotments in 1960. This also resulted in the poor colonists, who had been settled under the scheme of “advanced alienation,” being forced to undergo unnecessary and avoidable hardship.

(ii) Since funds were borrowed by Government at 6% interest, the additional cost of interest due to the avoidable delay of 10 years in respect of the above work was almost equal to the cost of the entire work itself! (Ibid. p.125).

(iii) To this loss has to be added the opportunity-cost of the value of the crops foregone during the avoidable 10 years of non-completion.

(iv) There were further avoidable delays caused by poor planning. For example, dam construction would take eight years while another two years were set apart for completing the channels, making 10 years in all, whereas they could have both been completed together in eight years, if programmed appropriately.

Highland Allotments

A major reason for the economic losses of the colonization schemes was that they left the highland allotments, which accounted for 40% of all colony lands before 1970 virtually uncultivated. In the latest schemes, the highland lot has been reduced to minimal proportions probably as a means of reducing the impact of this problem, but not of solving it.

1. From the inception of colonizaton, the highland allotments were completely cleared and stumped but then left, for the most part, uncultivated, thus causing tremendous soil erosion, the greatest hazard to Dry Zone cultivation.

2. It is essential to find a cultivation plan for the highland allotments. If the augmentation of the irrigation reservoirs discussed above is possible, the additional irrigation so provided should be directed only to the highlands for water-sparing, high-value crops. In order to prevent wastage and stealing of water meant for the highland allotments, it is desirable that the water be conveyed by pipe, with outlets for every group of four contiguous home-lots, with their owners made responsible for rotational, controlled and socially monitored water-use.

3. Wherever such tank augmentation is not feasible, lift-irrigation to the highland lots should be considered. This could be done, for example, by water being pumped directly from the reservoir to a higher level concrete trough, from which piped water could be provided for limited irrigation to the highland lots. Such lift-irrigation would need to be strictly controlled, with a fixed quantity of water allowed to each group of (say) four contiguous colonists only during the dry season, and only for water-sparing and high-value crops.

4. The introduction of tree-crops under irrigation should be considered. This idea was dismissed out of hand by the Minister of Agriculture and Lands in 1960. But piped water could be provided for trees by “deficit irrigation” for only the four months of the dry season each year by means of limited drip irrigation. Cheap means of doing so have now been devised in other Asian countries, using cheap plastic tubes and bottles to provide limited irrigation for different types of fruit trees. Such irrigation could also be used for fodder trees to support a livestock industry.

It is well recognized that there will be technical difficulties in providing irrigation to the highlands, especially because of problems of conveyance, water pilferage, water wastage and water sharing. But these are problems that need to be surmounted and not to be avoided, if the colonization schemes are to realize their full potential. While the above proposals are meant as a means of rehabilitating the highlands of the older schemes, they could also be considered for inclusion in future schemes.

Paddy Land Allotments

The paddy lands, on which all irrigation was lavished, were seriously underutilized in 1960 for the following reasons. It is encouraging to note that all the needed lines of development set out in the Short-Term Implementation Programme of 1962 have subsequently been followed up, thus enabling at least the paddy lands of the colonization schemes to live up to their great potential.

1. The yield obtained on paddy lands in colonization schemes in 1960 was only 40 bushels per acre, with only one crop per year, making these schemes completely uneconomic. On the other hand, there has been outstanding progress made in subsequent recent years, which has led to a doubling/trebling of yields, enabling these schemes to realize their real potential.

2. Although two crops of paddy were supposed to be cultivated per year, not even 8% of the paddy lands in all colonization schemes were cultivated for the second crop in 1960, due to wastage of water in the Maha season. For example, the Maha Illupalama Farm was achieving 30% higher yields for the Maha crop in 1960 than the adjoining colony fields (fed from the same tank) while using only one-third the water used by the colonists.

3. This wastage of water was allowed to continue by the colonists and by the Government, which had no conscious policy to improve water management. This was stressed (for the first time in a policy document in Sri Lanka) in the Short-Term Implementation Programme (1962).

4. Fortunately, there has been tremendous improvement since the 1980s, firstly, in terms of the Government’s active promotion of irrigation water management, and secondly, in terms of the farmers’ own action in saving water for the Yala season. This has enabled the achievement of much higher cropping intensities, multiplying positively the economics of these schemes.

5. All past calculations (prior to 1960) of the profitability of irrigated colonization schemes painted a rosy picture of their economics, based on incorrect calculations which obscured the sad reality. For example, on the cost side, the interest paid on capital was never counted, although the long delays in completion made these costs outrageous. On the benefit side, a yield of 40 bushels per acre was calculated for the paddy lands at a price of Rs. 12 per bushel, which was the price paid by the Government to farmers under the Gauranteed Price Scheme (GPS). But in practice, the real cost of rice in 1960 was around Rs.6.35 per bushel, the price at which the Government imported rice from abroad. Hence, the benefits from the paddy lands should be valued at Rs. 6.35 per bushel (the real economic value) and not at the subsidized price of Rs. 12 per bushel. In fact, the extra Rs. 6 (approx.) was actually a government subsidy to the farmer, which should be counted as a cost rather than as a benefit. This lopsided accounting masked the true economic reality of these schemes.

Conclusion

After initial antagonism and after two weeks of intensive briefing, the Prime Minister (Madam Bandaranaike) finally accepted the policy arguments set out above. Hence a policy paper was prepared at her bidding, setting out the facts, diagnosing the problems and outlining possible options to overcome them. This paper was circulated to the Hon. Minister of Agriculture and Lands, his two Permanent Secretaries and Heads of Departments, inviting them to a meeting chaired by the Hon. Prime Minister. At the meeting, the Hon. Minister, Mr. C.P de Silva, initially contested the facts and figures as well as the analysis and policy conclusions presented, but ultimately had to agree to them, since the figures had been provided by his own Ministry. He also, albeit reluctantly, agreed to them being written in the Short-Term Implementation Programme of 1962. However, about ten days later, the Hon. Minister went back on this agreement and threatened to cross the floor to the UNP with 14 of his supporters, if the policy changes agreed to at this meeting were to be enforced. Since this would have caused the collapse of the Government, the Hon. Prime Minister decided that she could not carry out the agreed changes at that time, and personally conveyed her regrets to me. However, at this point, realizing that all my work had come to naught, I decided to leave the country. I applied to FAO for a job and left the country in 1962. But ultimately when the Prime Minister, Mrs. Bandaranaike, tried to move Mr. C.P. De Silva from the Ministry of Agriculture about two years later in order to carry out the above policy changes, he crossed the floor with his supporters to join the UNP, thus bringing about the collapse of the Bandaranaike government. In return, he was rewarded by the new UNP Government with the portfolio of Agriculture and Lands, so that he could continue to do more of the same. Thus, despite the wake-up call of 1962, the irrigated colonization schemes went unquestioned and relatively untouched for about 20 years more, with many of the above defects being carried into the Mahaweli Scheme too in its early years.

I must repeat again that from a long-term economic point of view, these colonization policies were necessary for the country’s long-term economic future. The problem was that their planning and implementation left much to be desired. Moreover, this one-sided pursuit of colonization as a solution to our food and agricultural development problems led to a 10 year delay in reaping the fruits of the Green Revolution which were already within our grasp by 1960.2 Nor is this meant as an attack on Mr. C.P. de Silva, who dedicated his career in public service and political life to opening up the Dry Zone.

He is well known as an honest, devoted and dedicated public servant and as a loyal, generous and humane boss. It is, however, true that as Minister of Agriculture and Lands for around eight to ten years, he presided over a land development policy to the detriment of an agricultural development policy, while the former was characterized by poor planning and implementation, low physical output, high costs and uneconomic returns.

Fortunately, with the advent of the high yielding varieties and improved water management (both of which were stressed as early as 1962 in the Short-Term Implementation Programme), continuing improvements have been made over time, till now the colonization schemes are beginning to fulfill their purpose and their promise – although they have taken over 40 years (from 1960) in order to do so. Unfortunately our developers of the Dry Zone have left unheeded King Parakramabahu’s guiding principle of not letting a single drop of water (or land) go waste without first being adequately utilized by man. Our policy-makers (from 1950 – 1980) were thus content to leave 40% of the land in the colonization schemes (the highland allotments) unutilized, while also allowing 50% of the land and water (meant for the Yala cultivation) to go waste. Wise King Parakrama Bahu would not have been pleased to see what was being done in his name!

1 Unfortunately in one of two printed versions of this Programme, a misprinted footnote, instead of indicating that the paddy yield was calculated at 42 bushels, mistakenly indicated (inverting the numerals) that a yield of 24 bushels per acre was used.

2 The Indian Five Year Plan of 1965 is given the credit of heralding the Green Revolution, although in practice, it was preceded in this respect by the Short-Term Implementation Programme of Sri Lanka in 1962.